It is well established that higher socioeconomic status is associated with better health and lower mortality. Nevertheless, empirical results for historical periods are mixed and the understanding behind mortality patterns remain scarce. Most studies use a period perspective neglecting potential cohort effects in explaining socioeconomic differences. The aim of this paper is to study socioeconomic differences in adult mortality and analyze the effects of age, period, and cohort separately. We use longitudinal micro-level register data from both historical and contemporary sources linked together about southern Sweden. While the historical source of information is the Scanian Economic Demographic Database, data from 1969 onwards are taken from Swedish population registers. Both sources provide socioeconomic and demographic variables at the individual level. Preliminary results highlight the importance of cohort effects over period and age effects, and highlight a reverse socioeconomic gradient for cohorts in the second part of the nineteenth century. Differences in behavior or life style factors between social classes, in particular with respect to smoking, could explain such pattern.